Transcript Annular Comb Gate

HELIX: A Versatile Plasma Composition Spectrometer
Robert Sheldon, Dennis Gallagher, Oleg Vaisberg, Mark Adrian, Wheaton College/NSSTC/MSFC
Abstract
Composition instruments have made themselves indispensable in understanding the
complex plasma environment of both magnetospheres and the solar wind. The
combination of energy, mass and charge state effectively paint plasma constituents
and provide valuable information on sources, acceleration, transport and decay of
ions. However, future mission opportunities are likely to put severe constraints on
present instrument designs, requiring a radical rethinking of mass spectrometry
techniques.
We present a plasma composition spectrometer prototype based on flight-proven
time-of-flight technology with a novel helical design that greatly extends the flight
path length, and hence the resolution of the compact spectrometer. In addition, the
design enables very high geometric factors and duty cycles, making this instrument
highly sensitive for both rare species and fast sampling. The prototype has
demonstrated M/dM=50 at 2keV for ions and molecules, with the promise of M/dM
> 3000, all in a package that is estimated to be 1kg, 1 liter, and consume 1 Watt.
Scaling this package down to a “standard” resolution instrument with M/dM=100,
reduces these sizes to 100g 125cc spectrometer ideal for multiple space probes or a
Pluto flyby.
In this poster, we will try to trace the history of Space Mass Spectrometry, indicate
where we are, and what the future holds. It is our contention that we are at the
threshold of high resolution space spectrometers based on TOF technology, that will
make the quantitative measurements needed for 21st century models, both solar
plasma origins and of biological origins.
The Challenge of Space Mass Spectrometers vs Lab
Space
Laboratory
•Light weight: ~kg, 10-100 kg
•Small size: ~liter 10-1000 L
•Low power: ~Watt 100-1000 W
•Vibration limit: 10g @ 100Hz Fragile!
•UV rejection: 10 UV/s < 1/s
•Radiation hard: 10-100 kRad < 1 kRad
•Autonomous: MTBF years < week
Limitations of Space MS (other than resolution)
•Ion Traps/Faraday cups: M/dM ~2 best case
- Highly integrating measurement, adding all species and directions together. It does fine if moments
of the distribution function are sufficient information, rather than f itself.
•E/Q: Can’t separate He
++/C6+/O8+,
dE/E~5%
- Unfortunately much of the solar wind has the same E/q range, so that little composition
information is possible. Also, most analyzers are narrowband E/q filters that must be scanned,
with energy resolution inversely proportional to duty cycle. 5% dE/E is about as good as it
gets.
•Mag: E<25 keV, heavy magnets,M/dM~10
- Magnets separate on rigidity, or M/q, so they do a better job than E/q filters on composition.
However, the weight of magnets (or power) limits the highest energies to ~20keV/q, which
isn’t high enough to see the heavies in the solar wind. Again, resolution is inversely
proportional to sensitivity, giving a practical resolution limit of ~10.
•Linear TOF w/C-foil: E<2 keV/nuc, Energy straggle in C-foil limits M/dM~15
- Linear TOF measure velocity, which can be combined with an energy measurement to
determine E, m, and q separately. However, C-foils need > 2keV/nuc to work effectively,
which typically requires 10’s of kV post-acceleration before the C-foil. This reduces the TOF,
so that relative timing uncertainty grows. Thus there is a tradeoff between dE/E energy
straggle from the foil, and dT/T timing uncertainty that give a “sweet spot” for postacceleration around M/dM~15.
•Isochronous TOF: M/dM < 100; efficiency<10%,;dynamic range~200; E ~ 2keV/nuc--30keV
1961
•Electrostatic E/Q sorting (e.g. Ulysses) 2-4 1970
•Wien filters (e.g.ISEE)
2-6
1972
•Magnetic sector (e.g. POLAR/TIMAS) ~7 1980
•Linear Time-of-Flight (e.g.AMPTE/CHEM)~10 1984
•Isochronous TOF (e.g.WIND/SMS)
~100
1994
•Gated TOF (e.g. Rosetta/Rosina) ~2000 2011
•Helical TOF (??)
~3000 20??
Science attainable at various Resolutions
•1961 : ~2 (separation of H & He)
–pressure, velocity, density of solar wind plasmas
•1970 : ~10 (electrostatic: H,He,C,O,Fe)
–temperature, composition (during cold plasma conditions)
C/N/O, Si/S/Mg)
–fast/slow solar wind, coronal acceleration mechanisms
•1994 : ~100 (ITOF: Ne,Mg,Fe isotopes)
•
•Temporal Hadamard Transforms
-A paper in ROSI [2000] showed how this can be applied to TOF systems. Instead of the spatial
domain, the time domain can have a coded mask. Thus 50% open gates can be used to give the same
resolution with maximal sensitivity. The cost is in S/N ratio, and subtleties of the “kind” of noise in
the system. Signal independent noise can be reduced, but signal dependent noise (such as scattered
ions or neutrals) can not be addressed this way.
•Spatio-Temporal Transforms
-We have not done the calculation, but it seems feasible that space and time can be deconvolved
separately, leading to imaging TOF MS. This permits magnetic sector MS to be simultaneous with TOF
MS at 50% spatial and 50% temporal duty cycle. The advantage is to permit one technique to provide,
say, 4 low bits of resolution, and the other 4 high bits of resolution in a very compact MS.
In-line HELIX TOF Mass Spectrometer Concept
- protosolar nebula composition, solar mixing ratios
200?: ~3000 (HELIX: CO/N2 molecules, 16/17/18O ratios)
Shocks transport, magnetic holes, cosmochemistry, biomolecules
Why aren’t dynamic range & sensitivity important?
Laboratory measurements often are looking for rare species and
need lots of sensitivity, ppm to ppb. Space plasmas come preionized and pre-energized, and thus are a fixed quantity. Since
space plasma densities are orders of magnitude below laboratory
“ionizers”, the critical factor for space MS is geometric factor:
how big an opening is available for collecting space plasmas.
Since UV photons come through the same opening, generally
the limitation for space is determined by the collimator/ light
trap.
with 0.1%, dE/E ~ 1 %, whereas TOF can be determined with ppm making it the method of
choice.
•For asynchronous, low countrate conditions, nothing beats the efficiency of C-foils in TOF
measurements. All asynchronous systems saturate, however, giving C-foils an effective upper
limit on countrate around 1-100kHz.
Potential Housing
Ions
N2+
C+
O+ OH+
+
O
17
+
N
Collimater
Annular
MCP
Wire Comb
TOF Gate
Hyperbolic Mirror
Vg
The Chopper
•A chopper is a fixed time signal, a synchronous measurement
-used extensively since at least Fizeau, ~1800
-used in laboratory/space TOF, e.g. ROSETTA TOF
•For ions, it need not have moving parts, using electrostatics.
-ENA is a different story but may have a MEMS solution
•It trades efficiency vs. resolution. The narrower the gate, the higher the resolution, but
the lower the sensitivity.
- Ambiguities are generated for multiple packets.
-This can be partially solved with Hadamard transforms.
-Alternatively, combining magnetic sector & chopped TOF can remove the ambiguity with
better noise tolerance.
-Position sensing also can be used to remove ambiguities, giving this design a number of
alternative paths toward higher sensitivities.
The greatest challenge is to make the chopper fast enough for the energetic plasma ions
of the solar wind, ~ 1-3 keV/nuc.
- 1us gate widths are standard, but 10-100ns gatewidths are needed for >10keV ions
- Recent space instruments are beginning to achieve these higher frequencies without a
huge penalty in power. One has to be careful, however, of radiating too much EMI.
TOF Resolution and UV Rejection
•Linear TOF are limited by total flight time
–50 cm is perhaps a practical limit to length, which for 1 keV H+ => ~1 us. This is about
the same time as a gate pulse, which means the uncertainty in time is ~ 50%.
–Electrostatic mirrors (e.g. ROSETTA), can double or quadruple TOF, but not 10x, 100x.
•Linear TOF doesn’t reject neutrals & photons
–separate UV traps have to be installed adding further restrictions to the geometric factor.
•Long TOF & UV rejection are both possible with a helical path.
-This may be the best of both worlds.
Therefore, with a long, helical path, TOF may be increased by 10-100x, resulting in 10100 improvement in dT/T, and resolution, while simultaneously excluding photons.
O2+
Ar +
0.1 µs
Time Delay (s)
(energy straggle) the efficiency (most particles exit as neutrals), and dynamic range. But no
replacement for C-foils has ever been found that is any better in charge state, energy straggle,
or strength...
foilless systems must be developed. The most common alternative is a Chopper, a
synchronous start gate.
NO+
Ar ++
H2O+
•In addition, Carbon foils set a lower energy limit (~1keV/nuc), and limit the resolution
•THEREFORE to increase countrate, energy range and dynamic range for TOF technology,
CO2+
2 keV beam
Mirror voltage = 1050V
CO2 gas feed
Advantages of Time-of-Flight Technology
A Short History of Space MS vs Resolution (M/dM)
•1984 : ~15 (LTOF:
-Suppose one uses a pinhole camera to take gamma-ray pictures of galaxies. Clearly, two pinholes
give twice the sensitivity, but make for overlapping images. With some effort, we can deconvolve the
image achieving the same resolution at twice the sensitivity. Extrapolating then to a coded mask made
of 50% pinholes allows the maximal sensitivity which can still be deconvolved. This is standard
practice in high energy astrophysics.
Ions
•TOF has advantages over magnetic sector and energy selection since position is determined
2-4
•Spatial Hadamard Transforms
-The highest mass resolution yet flown (Rosina not yet operational), but pushing the limits of
C-foil technology. Uses a hyperbolic field to eliminate dependence of TOF on E, thus immune
to energy straggle in the foils. However, 80+% of ions exit the foil in a neutral state, and
become noise rather than signal. This limits the efficiency and the dynamic range. Postacceleration can partly mitigate this effect, but is limited by the deflection voltages available
in space to V<30kV. At these voltages, TOF is also severely reduced, with the same problems
as above. Nor is this voltage high enough to permit a larger efficiency with C-foils. Nothing
better than C-foils is likely to exist.
8
•Ion traps, Faraday cups (e.g.Imp 8)
•Is Resolution ALWAYS Inversely Proportional to Sensitivity?
Analysis
In the spectrum above, a mixture of gases was introduced through the leak
valve into a electron impact ionization source, and accelerated to 2 keV. The
Gate
Gate Closed
Gate Closed
beam was primarily carbon dioxide, but with some atmospheric and “P40”, a
Open
H+ Mode
Heavy Mode
mixture of argon and methane, in the system from previous runs. Notice that
without a C-foil, the molecular gases do not dissociate. Note also, that the
Collimator/Buncher
width of the peaks appears to depend on mass, whereas strict TOF gating
In this concept, ions are collimated through a cylindrical collimator that has baffles to exclude UV indicates that the width ought to be independent of mass. We interpret this
sunlight. This collimator can also work as an “ion buncher” which operates like “chirping” in radar, effect to be due to finite width of the comb-gate, interacting with the slower
moving molecular species.
to take a collection of ions entering in time period dt, and compress dt by spreading dE, such that
dt*dE=constant. Since our instrument is an isochronous instrument, dE has little effect on the TOF. Also note the near absence of background. This is remarkable, since our
Thus a buncher can improve the efficiency of a synchronous gate to nearly the same efficiency as
gating technique is spewing ions all over the chamber, and we are intercepting
an asynchronous C-foil.
only 5% or so of entering ions. We attribute this to very little scattering within
TOF
Annular Channel Plate
The annular multichannel plates permit the beam to pass through the gap, but on the return, they
detect the ions. The cylindrical electric field is maintained continuously through this region by
oversized outer cylinder kept at a higher potential, such that the field is smooth. Position sensing
anodes can be implemented if the dual-Hadamard system is employed, though all such
improvements do make the design more complicated. Likewise, should total energy, E (and hence,
q) be desired, the annular MCP can be replaced by a annular SSD with a cylindrical MCP (not
shown) used to record the stop time. Such complications are only necessary if E/q information
proves ambiguous, which is rarely the case when M/dM>1000.
the TOF system itself at pressures of about 2-3x10-6 Torr. With proper
collimation, the noise contribution from the deflected beam can be eliminated
giving an additional 10-100x improvement in our already excellent S/N ratio.
We expect the dynamic range of this instrument to be more than sufficient to
measure the 16/17/18Oxygen isotope ratios.
Conclusions
TOF has the promise to revolutionize space mass spectrometry. We have
achieved M/dM=50 without C-foils, and without magnets, for molecules as
Annular Comb Gate
heavy as carbon dioxide. The mass resolution is exactly as designed, with a
The modulation is effected with a comb gate, which has the advantage of very small capacitance
10us TOF and a 100ns gatewidth. (Note that T/dT=100 is equivalent to
and inductance, and therefore very fast switching times. Laboratory prototypes have been switched M/dM=50). We had expected to increase this mass resolution to several
at 80ns, and probably 10ns is easily obtained with a more advanced pulser. The gate operates with
hundred by lowering the energy, and thereby raising the TOF. Unfortunately,
alternating wires of high potential that deflect the ions into the walls of the TOF region (OFF state).
the
finite
thickness
of
our
comb-gate
prevented
this,
since
slower
moving
ions
When the voltages are grounded (ON state), the ions proceed through. On return of the reflected
interacted with the gate even when relatively far away. Heuristically, we were
ions, the OFF state deflect the ions onto the two annular detectors. Since deflection depends on
trying to chop the ion beam with a dull knife, and despite feeding the beam in
voltage, one can also have an OFF’ state that deflects ions away from the detectors. Should the
more slowly, could not achieve thinner slices. This necessitates a second
detectors be swamped by H+, one can put the comb gate into an OFF’ state to avoid saturation.
generation comb gate that is currently under development, but which should
The TOF region
have a 1/10 thickness compared to our current hand-wound gate. This
A patented reflection region confines the ions and provides a hyperbolic reflecting field that makes suggests that 10ns might be close to the limit of achievable gate widths.
Accordingly our second generation design will have a 100us flight time, for a
for an energy independent TOF. This is the heart of the device, and the part we have tested in our
first prototype shown above.
combined resolution of ½ (dT/T) = 5000.